Somebody convince me; UCCNC for $450AUD vs LinuxCNC for free

[beefy]

One other thing that attracted me was tool wear compensation. I won't need this until I have a lathe or mill, but if it's true, UCCNC does not have this and Linuxcnc does.

UCCNC does have cutter comp, which basically can do the same thing. Wear comp is usually just a separate value that's added to the cutter comp.

Thanks Gerry.

In UCCNC, do you know if you can pause a running gcode program, go to the tool table and modify a tool's diameter, then when you continue running the gcode the new tool diameter setting will be used.
I had a play in simulation but when I pressed feedhold and went to the tool table, it was greyed out and could not be modified.

I've been Googling for wear compensation in Linuxcnc but so far only came across people asking for it, so now I'm doubting is it even has it.

[cncdrive]

Tool wear compensation is basically G41/G42 with the tool radius + tool wear value.
In the UCCNC there is no tool wear, but you could hack it with adding your own DROs for tool radius and tool wear and applying the sum to the tool dia DRO used for the G41/G42.

Changing the tool radius for compensation while running and pausing code is not possible though, you have to stop the code to do that and you have to reapply the G41/G42 to apply the new value.

[beefy]

Thanks Balazs,

I'm guessing the gcode is read from the beginning and the moves calculated and put into a buffer or something, and that's why you can't simply change things in the middle of a file run.

[hmnijp]

One other thing that attracted me was tool wear compensation. I won't need this until I have a lathe or mill, but if it's true, UCCNC does not have this and Linuxcnc does.

UCCNC does have cutter comp, which basically can do the same thing. Wear comp is usually just a separate value that's added to the cutter comp.

G41/G42 cannot be used in UCCNC - there are a lot of errors with the wrong approach and departure on simple path, this causes damage to the parts.

Compensation must be switched on strictly on the section of the path where the code G41/42 is written - on this line the transition from the path to the equidistant is drawn.

But in practice in UCCNC, the offset starts to be used on the line earlier or later than necessary - this causes incorrect path errors. The further construction of the equidistant line is correct.
It is not possible to configure a post processor to work with this. It is necessary to correct these errors in the program code, otherwise there is no point in declaring support for diameter compensation.

- if someone wants to convince me otherwise - apparently he does not use it much. I described it in detail in this thread http://cncdrive.com/forum/viewtopic.php?f=3&t=3003 and attached screenshots. Since that time, nothing has changed.

Also, for a turning configuration, not only tool length compensation is needed (G41 Z), but also compensation for ALL axes (G43 X Y Z) and additional lathe tool orientation.
It's also nice to have separate columns for wear and radius in the tool table, and switch between using them.

LinuxCNC does not have such problems - this is its plus if you need a serious system more like an industrial one. But if you do not need it and want a simple friendly system for router - your choice is UCCNC

[cncdrive]

I just fixed this problem (except when starting with G19G3 arc, that is currently not handled by UCCNC, I mean parts in other than in G17 with compensation, will try to implement that later).

[hmnijp]

I just fixed this problem (except when starting with G19G3 arc, that is currently not handled by UCCNC, I mean parts in other than in G17 with compensation, will try to implement that later).

this is really good news!
Thank you for your response, I will wait for the new version!

[cncdrive]

No problem. We still working on the lathe support development, but when that will be finished then we will make a new release which will include this fix too and many new developments too.
BTW, do you have any other g-codes which I can test the G41/G42? I mean which codes works wrong with the current release with the G40/41/42?
If you have then please upload them to ket me test them.

[muzzlitebeer]

Well, it's done.

I'm now sitting next to a CNC router connected to my PC via an AXBB-E controller.

Anybody who says they got this up and running in half an hour is either lying, or they've done it before. It wasn't easy.

Not reading the proper manual didn't help. I was following the guide for UCCNC software instead of the AXBB-E manual.

Ah well... I got there in the end.

Thank you for all the encouragement. Bye Bye LinuxCNC. Bye Bye ancient parallel port desktop PC.

[Authority924]

LinuxCNC looks pretty good to me, well I do not have fully understand how it works but at the beginning it looks pretty good

[beefy]

Bye Bye LinuxCNC. Bye Bye ancient parallel port desktop PC.

Seeing as this thread is sort of UCCNC vs Linuxcnc, then just to be fair and not make it seem like you must use an old PC, many Linuxcnc users do not use an ancient parallel port PC. That was totally your choice, not something forced on you.

Instead they opt for one of the many high quality Mesa motion control cards, which connect via ethernet. That is very similar to UCCNC.

Plus the motion controller board I use in my Linuxcnc plasma setup does not even require a breakout board, as the opto isolation and signal buffering is all done on the motion controller board. It's effectively a motion controller board and breakout board in one, which saves money and simplifies wiring. I think the AXBB board from Cnc drive is similar.

So modern PC plus one motion controller board, and no separate breakout board. Not free but not expensive either, and top quality.